arXiv subjects

Partha Sarathi Goswami

Publications and source records attributed to Partha Sarathi Goswami.

6 recordsLinked to original sources

arXiv2023

Standing waves and jets on a sessile, incompressible bubble

We show numerically that large amplitude, \textit{shape deformations}, imposed on a spherical-cap, incompressible, sessile gas bubble pinned on a rigid wall can produce a sharp, wall-directed jet. For such a bubble filled with a permanent gas, the temporal spectrum for surface-tension driven, linearised perturbations has been studied recently in \citet{ding2022oscillations} in the potential flow limit. We reformulate this as an initial-value problem. Linear theory is validated by distorting the shape of the pinned, spherical cap employing eigenmodes obtained theoretically, as the initial perturbation for our numerical simulations. It is seen that linearised predictions show good agreement with nonlinear simulations at small distortion amplitude producing standing waves. Beyond the linear regime, we observe the formation of a dimple followed by a slender, wall-directed jet analogous to similar jets observed in other geometries from collapsing wave troughs\cite{farsoiya2017axisymmetric,kayal2022dimples}. This jet can eject with an instantaneous velocity exceeding nearly twenty times that predicted by linear theory. By projecting the shape of the bubble surface around the time instant of jet ejection, into the linearised eigenspectrum we show that the jet ejection coincides with the nonlinear spreading of energy into a large number of eigenmodes. We demonstrate that the velocity-field associated with the dimple plays a crucial role in evolving it into a jet and without which, the jet does not form. Our inferences also complement well-known results of \citet{naude1961mechanism} and \citet{plesset1971collapse} demonstrating that wall-directed jets can be generated from \textit{volume preserving}, shape deformation of a pinned bubble.

physics.flu-dyn↗

arXiv2023

Effect of channel dimensions and Reynolds numbers on the turbulence modulation for particle-laden turbulent channel flows

The addition of particles to turbulent flows changes the underlying mechanism of turbulence and leads to turbulence modulation. Different temporal and spatial scales for both phases make it challenging to understand turbulence modulation via one parameter. The important parameters are particle Stokes number, mass loading, particle Reynolds number, fluid bulk Reynolds number, etc., that act together and affect the fluid phase turbulence intensities. In the present study, we have carried out the large eddy simulations for different system sizes (2δ/dp = 54, 81, and 117) and fluid bulk Reynolds numbers (Re_b = 5600 and 13750) to quantify the extent of turbulence attenuation. Here, δ is the half-channel width, dp is the particle diameter, and Re_b is the fluid Reynolds number based on the fluid bulk velocity and channel width. The point particles are tracked with the Lagrangian approach. The scaling analysis of the feedback force shows that system size and fluid bulk Reynolds number are the two crucial parameters that affect the turbulence modulation more significantly than the other. The streamwise turbulent structures are observed to become lengthier and fewer with an increase in system size for the same volume fraction and fixed bulk Reynolds number. However, the streamwise high-speed streaks are smaller, thinner, and closely spaced for higher Reynolds numbers than the lower ones for the same volume fraction. In particle statistics, it is observed that the scaled particle fluctuations increase with the increase in system size while keeping the Reynolds number fixed. However, the scaled particle fluctuations decrease with the increase in fluid bulk Reynolds number for the same volume fraction and fixed system size. The present study highlights the scaling issue for designing industrial equipment for particle-laden turbulent flows.

physics.flu-dyn↗

arXiv2022

Dynamics of particle-laden turbulent Couette flow. Part2: Modified fluctuating force model (M-FFS)

Two-way coupled DNS simulation of particle-laden turbulent Couette-flow [1], in the volume fraction regime $ϕ>10^{-4}$, showed a discontinuous decrease of turbulence intensity beyond a critical volume fraction $ϕ_{cr}\sim7.875\times10^{-4}$. Due to the presence of high inertial particles, the drastic reduction of shear production of turbulence is found to be the main cause for the discontinuous attenuation of turbulence. In this article, particle-phase statistics is explored. The two-way coupled DNS reveal that the mean-square velocity profiles in cross-stream (y) and span-wise (z) directions are flat and increase with $ϕ$ as the higher frequency of collision helps in transferring streamwise momentum to span-wise and wall-normal directions. Whereas, streamwise fluctuations decrease and tend become flatter with increase in loading. In the regime with $ϕ>ϕ_{cr}$, the particle velocity fluctuations drive the fluid phase velocity fluctuations. Additionally it is observed that one-way coupled DNS and Fluctuating Force Simulation (FFS) [2] are capable to predict the particle phase statistics with reasonable accuracy in the regime $ϕ<ϕ_{cr}$ where wall-particle collision time and inter-particle collision time is lesser than viscous relaxation time of the particles. For, $ϕ>ϕ_{cr}$, a significant error in the prediction from one-way coupled DNS and FFS is observed due to the limitation of FFS in capturing the turbulence attenuation and the change in mean fluid velocity profile. A modified FFS model (M-FFS) is successfully developed in this article with modified mean fluid velocity profile and zero-diffusivity.

physics.flu-dyn↗

arXiv2022

Dynamics of particle-laden turbulent Couette flow. Part1: Turbulence modulation by inertial particles

In particle-laden turbulent flows the turbulence in carrier fluid phase gets affected by the dispersed particle phase for volume fraction above $10^{-4}$ and hence reverse coupling or two-way coupling becomes relevant in that volume fraction regime. In a recent study by Muramulla $et.al.^1$, a discontinuous decrease of turbulence intensity is observed in a vertical particle-laden turbulent channel-flow for a critical volume fraction O($10^{-3}$). The collapse of turbulent intensity is found out to be a result of catastrophic reduction of turbulent energy production rate. Mechanistically, particle-fluid coupling in particle-laden turbulent Couette-flow differs from that in a closed channel flow. In this article, the turbulence modulation in Couette-flow by inertial particles is explored through two-way coupled DNS where particle volume fraction ($ϕ$) is varied from $1.75\times10^{-4}$ to $1.05\times10^{-3}$ and Reynolds Number based on half-channel width ($δ$) and wall velocity ($U$) ($Re_δ$) is $750$. The particles are heavy point particles with $St\sim367$ based on fluid integral time-scale represented by $δ/U$. A discontinuous decrease of fluid turbulence intensity, mean square velocity and Reynolds stress is observed beyond a critical volume fraction $ϕ_{cr}\sim7.875\times10^{-4}$. The drastic reduction of shear production of turbulence and in turn the reduction of viscous dissipation of turbulent kinetic energy are two important phenomena for the occurrence of discontinuous transition similar to channelflow. The step-wise particle injection and step-wise removal study confirms that it is the presence of particles which is majorly behind this discontinuous transition.

physics.flu-dyn↗

arXiv2022

Effect of rough wall on drag, lift, and torque on an ellipsoidal particle in a linear shear flow

The present study provides a detailed description of the forces on an ellipsoidal particle in the vicinity of the rough wall. Three-dimensional numerical simulations are performed using body-fitted mesh to estimate the drag, lift, and torque coefficients. A large number of simulations are conducted %(approximately 2400) over a range of parameters such as shear Reynolds number ($10 \le Re_s \le 100$), orientation angle ($0\leθ\le 180$), and wall-particle separation distance ($0.1\leδ\le2.0$) to get a comprehensive description of variation of the above coefficients. Using the simulation results, we develop the correlations for the drag and lift coefficients to describe the effect of rough wall, inclination angles, and particle Reynolds numbers on the hydrodynamic coefficients. The proposed correlations can be used for two phase flow simulation using Eulerian-Lagrangian framework.

physics.flu-dyn↗

arXiv2021

A Statistical Analysis Towards Modelling the Fluctuating Torque on Particles in Particle-laden Turbulent Shear Flow

Dynamics of the particle phase in a particle laden turbulent flow is highly influenced by the fluctuating velocity and vorticity field of the fluid phase. The present work mainly focuses on exploring the possibility of applying a Langevin type of random torque model to predict the rotational dynamics of the particle phase. Towards this objective, direct numerical simulations (DNS) have been carried out for particle laden turbulent shear flow with Reynolds number, $Re_δ=750$ in presence of sub-Kolmogorov sized inertial particles (Stokes number >>1). The inter-particle and wall-particle interactions have also been considered to be elastic. From the particle equation of rotational motion, we arrive at the expression where the fluctuating angular acceleration fluctuation $α'_i$ of the particle is expressed as the ratio of a linear combination of fluctuating rotational velocities of particle $ω'_i$ and fluid angular velocity $Ω'_i$ to the particle rotational relaxation time $τ_r$. The analysis was done using p.d.f plots and Jensen-Shannon Divergence based method to assess the similarity of the particle net rotational acceleration distribution $f(α'_i)$, with (i) the distributions of particle acceleration component arising from fluctuating fluid angular velocity computed in the particle-Largrangian frame $f(Ω'_i/τ_r)_{pl}$, (ii) fluctuating particle angular velocity $f(ω'_i/τ_r)_{pl}$, and (iii) the fluid angular velocity $(Ω'_i/τ_r)_{e}$, computed in the fluid Eulerian grids. The analysis leads to the conclusion that $f(α'_i)$ can be modeled with a Gaussian white noise using a pre-estimated strength which can be calculated from the temporal correlation of $(Ω'_i/τ_r)_{e}$.

physics.flu-dyn↗